Yggdrasil v0.10.0 Release

Yggdrasil now has way more configuration options, the following options are all new:

--envoy-port uint32                        port by the envoy proxy to accept incoming connections (default 10000)
--health-address string                    yggdrasil health API listen address (default "0.0.0.0:8081")
--host-selection-retry-attempts int        Number of host selection retry attempts. Set to value >=0 to enable (default -1)
--max-ejection-percentage int32            maximal percentage of hosts ejected via outlier detection. Set to >=0 to activate outlier detection in envoy. (default -1)
--upstream-healthcheck-healthy uint32      number of successful healthchecks before the backend is considered healthy (default 3)
--upstream-healthcheck-interval duration   duration of the upstream health check interval (default 10s)
--upstream-healthcheck-timeout duration    timeout of the upstream healthchecks (default 5s)
--upstream-healthcheck-unhealthy uint32    number of failed healthchecks before the backend is considered unhealthy (default 3)
--upstream-port uint32                     port used to connect to the upstream ingresses (default 443)

New health check endpoint at /yggdrasil/status on your envoy nodes, allowing you to validate that they've started successfully and been configured by Yggdrasil.

Cluster names now no longer have any .s in them to stop strange prometheus metrics being generated see this Envoy Issue for more info.

Massive thanks to @hikhvar for contributing all of this!